Flow, Turbulence and Combustion最新文献

Direct numerical simulations (DNS) have been utilised to investigate the impact of different thermal wall boundary conditions on premixed V-flames interacting with walls in a turbulent channel flow configuration. Two boundary conditions are considered: isothermal walls, where the wall temperature is set either equal to the unburned mixture temperature or an elevated temperature, and adiabatic walls. An increase in wall temperature has been found to decrease the minimum flame quenching distance and increase the maximum wall heat flux magnitude. The analysis reveals notable differences in mean behaviours of the progress variable and non-dimensional temperature in response to thermal boundary conditions. At the upstream of the flame–wall interaction location, higher mean friction velocity values are observed for the case with elevated wall temperature compared to the other cases. However, during flame–wall interaction, friction velocity values decrease for isothermal walls but initially rise before decreasing for adiabatic walls, persisting at levels surpassing isothermal conditions. For all thermal wall boundary conditions, the mean scalar dissipation rates of the progress variable and non-dimensional temperature exhibit a decreasing trend towards the wall. Notably, in the case of isothermal wall boundary condition, a higher scalar dissipation rate for the non-dimensional temperature is observed in comparison to the scalar dissipation rate for the progress variable. Thermal boundary condition also has a significant impact on Reynolds stress components, turbulent kinetic energy, and dissipation rates, showing the highest magnitudes with isothermal case with elevated wall temperature and the lowest magnitude for the isothermal wall with unburned gas temperature. The findings of the current analysis suggest that thermal boundary conditions can potentially significantly affect trubulence closures in the context of Reynolds averaged Navier–Stokes simulations of premixed flame–wall interaction.

The present experimental study reports first observations of stability, blowout, and blowoff characteristics of ammonia–hydrogen–nitrogen fuel blend flames with varying volumetric ammonia fractions (({x}_{{NH}_{3}})) in a coaxial combustor. The ({x}_{{NH}_{3}}) is varied from 20 to 80%. For flames of ammonia fraction equal to 70% (({x}_{{NH}_{3}}=0.7)), three types of flame transitions are observed within fuel flow Reynolds number (({Re}_{f})) of 40–575 as a coflow Reynolds number (({Re}_{a})) is increased in steps. Initially, the coflow air remains laminar and ({Re}_{a}) is increased gradually from laminar to turbulent limit. Different flame stabilization modes are characterized as burner-attached and lifted flame. The flame extinction modes are classified as lifted-blowoff, attached-blowoff and attached-blowout types. These flame transitions and stabilization characteristics are shown to be similar to methane flames. However, the flame height and liftoff height are shown to be different. The flames of fuel blends with ammonia fraction less than or equal to 60% (({x}_{{NH}_{3}}le 0.6)) are shown to behave fundamentally different from that of flames with ({x}_{{NH}_{3}}>0.6) (and also methane flames). Specifically, within the tested ({Re}_{f}) range, only one type of flame transition is observed as ({Re}_{a}) is systematically varied in the former as compared to three types observed in the latter. Also, with a decrease in ammonia fraction (and a corresponding increase in hydrogen percentage), the liftoff limit, reattachment limit, and blowout limits all are observed to increase. The effect of ammonia composition on flame height and liftoff height is also elaborated. The present study also provides empirical correlations (particularly for the low power flames) for predicting blowout and blowoff limits in both lifted and attached conditions for ammonia-hydrogen–nitrogen fuel blend flames.

As the main by-product of converter steelmaking process, converter gas has significant potential for energy recovery due to its high calorific value. However, there is a significant risk of explosion during the recycling process. In order to ensure the process safety of converter gas recovery and achieve efficient energy utilization, it is necessary to study the process of CO deflagration in the tube and prevent it. This article combines experiments and numerical simulations to study the effects of obstacles inside tube, water content in the air, and the length of the smooth section on CO deflagration characteristics. The results show that the propagation characteristics of flames in the smooth section are related to the flow field and have periodicity. The length of the smooth section does not significantly affect the maximum deflagration pressure. During the propagation of flames in the obstacle section, the acceleration effect of each obstacle on the flame is similar, and the deflagration becomes more and more intense as the number of obstacles increases. The peak value is reached at the last obstacle, about 0.72 MPa, and the flame speed can reach 672 m/s. The water content in the air has a significant impact on the maximum deflagration pressure of CO, as H₂O triggers a series of chain branching reactions. When the water content increases to 0.39%, the maximum deflagration pressure reaches its peak. In terms of numerical simulation, the reliability of the open-source combustion solver XiFoam was verified. The combustion, transport, and thermodynamic property parameters for premixed gas of CO and humid air were provided using Cantera. Finally, in order to avoid the occurrence of deflagration during the converter gas recovery process, it is necessary to strictly control its moisture content.

The magnitude of the wrinkled flame surface area in turbulent premixed flames divided by its projection in the direction of flame propagation, known as the wrinkling factor, is a fundamental quantity for the purpose of analysis and modelling premixed combustion, for example, in flame surface density based modelling approaches. According to Damköhler’s hypothesis it is closely related to the turbulent burning velocity, an equally important measure of the overall burning rate of a wrinkled flame. Three-dimensional evaluation of the area of highly wrinkled flames remains difficult and experiments are often based on planar measurements. As a result of this, model development and calibration require an extension of 2D measurements to 3D data. Different relations between 2D and 3D wrinkling factors are known in literature and will be discussed in the present work using a variety of direct numerical simulation (DNS) databases combined with theoretical arguments. It is shown, based on an earlier analysis, that the isotropic distribution of the surface area weighted probability density function of the angle between the normal vectors on the measurement plane and the flame surface, provides a very simple relationship, stating that the ratio between 3D and 2D flame surface area is given by (4/pi ), which is found to be in excellent agreement with DNS data of statistically planar turbulent premixed flames.

This study presents the computation of the ignition probability map of a model gas turbine, investigated experimentally at CORIA laboratory, using Large Eddy Simulation (LES). The simulations leverage the recently proposed TFM-AMR-I model, which is based on the Thickened Flame Model (TFM) formalism and enables a full flame resolution (i.e. no thickening) of the flame kernel in the initial instants of ignition. LES simulations of ignition are performed for 14 spatial points distributed in the combustion chamber, with 6 repetitions for each in order to obtain a reasonable estimate of ignition probabilities. Probabilities are adequately predicted for most of the selected points, with a typical error of 30 (%). Nevertheless, the ignition probability is largely over-estimated at two locations where the mean diameter of liquid droplets is shown to be under-predicted, which may lead to too easy ignitions. Parametric variations show a satisfying robustness of the proposed approach with the two following key highlights: (i) the initial full flame resolution made possible by TFM-AMR-I is necessary, as an abrupt initial thickening leads to an artificial extinction; (ii) a correction of the over-sensitivity of the thickened flame to stretch, recently proposed in the literature, is necessary to predict ignition accurately.

In combustion theory, flames are usually described in terms of the dynamics of iso-surfaces of a specific scalar. The flame displacement speed is then introduced as a local variable quantifying the progression of these iso-surfaces relative to the flow field. While formally defined as a scalar, the physical meaning of this quantity allows relating it with a vector pointing along the normal direction of the scalar iso-surface. In this work, this one-dimensional concept is extended by the introduction of a generalized flame displacement velocity vector, which is associated with the dynamics of iso-surfaces of two generic scalars, (alpha ) and (beta ). It is then shown how a new flamelet paradigm can be built around this velocity vector, which leads to (i) an alternative procedure for the derivation of general flamelet equations, which is much simpler and more direct than the ones currently available in the literature, (ii) a very compact set of two-dimensional flamelet equations for the conditioning scalar gradients, (g_{alpha } = |nabla alpha |) and (g_{beta } = |nabla beta |), which comprise several effects in few terms directly related to the projections of the generalized flame displacement velocity, and (iii) the possibility of characterizing different composition space coordinate systems through the same generalized flame displacement velocity. The proposed framework is discussed in the context of partially premixed combustion, emphasizing how its adoption can contribute to both the further development of 2D flamelet theory and its coupling with CFD codes.

The effect of inlet mixture stratification was investigated in propane/air and prevaporised n-heptane/air flames stabilized in the near wake region of a bluff-body burner. The employed axisymmetric burner can sustain flame anchoring at global equivalence ratio values in the range of 0.09 ÷ 0.1 independently of fuel type and permits the variation of fuel concentration along the radial direction. Three distinct stratification gradients were studied for the two fuels considered; One burning from rich to lean, one burning from stoichiometric to lean and one burning from stronger lean to weaker lean mixtures. Particle Image Velocimetry, Mie scattering and OH ∗ and CH* Chemiluminescence were used to investigate flame stabilization characteristics of the two fuels and three stratification gradients, while Fourier – Transform Infrared Spectroscopy was performed to assess the equivalence ratio disposition under non-reacting conditions in the near wake region. 2D hydrodynamic strain rates, Damköhler (Da) and Karlovitz (Ka) numbers and flame brush thickness distributions were estimated and analyzed to elucidate the effects of turbulence, mixture composition and fuel type on the investigated flames. Also, the characteristic size of the reacting fluid pockets was assessed using a two-point sample autocorrelation methodology on the OH* chemiluminescence images. Results suggest that supplying the vicinity of the anchoring region with lean peak equivalence ratio mixtures with Lewis numbers greater than unity reduces the flame’s resistance to strain, while supplying it with rich peak equivalence ratio mixtures of Lewis number ≈1, independently of fuel type, favors resistance to strain, suggesting a connection with preferential diffusion effects.

Recent advances in combustion modelling for Large Eddy Simulation (LES) have increasingly utilised lower-dimensional manifolds, such as Flamelet Generated Manifolds and Flamelet/Progress Variable methods, due to their computational efficiency. These methods typically rely on one-dimensional representations of flame structures, often assuming premixed or non-premixed configurations. However, practical combustion devices frequently operate under partially-premixed conditions and present challenges due to mixture inhomogeneities and complex flow features. The Linear Eddy Model (LEM) offers an alternative by directly simulating turbulence-chemistry interactions without presuming specific flame structures. However, traditional LES-LEM approaches are computationally quite expensive due to the need for resolved LEM domains to be embedded in every LES cell.The authors developed the Super-Grid LEM (SG-LEM) method (Comb. Theor. Model.  28, 2024) to address these computational challenges by coarse-graining the LES mesh and embedding individual LEM domains within clusters of LES cells. This study evaluates SG-LEM in the context of the Multi-Regime Burner (MRB) introduced by Butz et al. (Combust. Flame, 210, 2019), which features both premixed and non-premixed flame characteristics. SG-LEM simulations of the MRB case demonstrate the method’s sensitivity to clustering parameters, with flow-aligned clusters significantly improving flame stability. LEM domains on the super-grid were able to represent the MRB flame topology while LES radial profiles including velocity, mixture fraction, temperature, and ({textrm{CO}}) mass fraction, were validated against experimental data and also reference simulations using standard combustion closures. The work also investigates discrepancies in CO profiles using conditional statistics and stand-alone LEM simulations. Finally, the work identifies areas of improvement for the SG-LEM framework, in particular relating to cluster generation, and (advective and diffusive) mass exchange between neighbouring LEM domains, as well as possible solutions for future SG-LEM implementations which could improve the model’s predictive capability.

Hybrid RANS/LES methods can produce more reliable results than RANS with a reasonable computational cost. Thus, they have the potential to become the next workhorse in the industry. However, in continuous approaches, whether or not they depend on the grid step explicitly, the ability of the model to switch to a well-resolved LES depends on the mesh generated by the user, such that the results are user-dependent. The present paper proposes a self-adaptive strategy, in which the RANS and LES zones are determined using physical criteria, in order to mitigate the user influence. Starting from an initial RANS computation, successive HTLES are carried out and the mesh is refined according to the criteria. To demonstrate the feasibility of this strategy, the method is applied to the backward-facing step case with the Hybrid Temporal Large Eddy Simulation (HTLES) approach, but is suitable for any other hybrid approach. The results obtained show that the method reaches a fixed point after only a few simulations and significantly improves the predictions when compared to RANS, with no intervention from the user. Even though the process is still a long way from being applicable to a wide range of turbulent flows, this paper is a demonstrator of the applicability of this self-adaptive strategy.

Mixing models for multiple mapping conditioning (MMC) methods are revisited as some details of their implementation have not yet been assessed. We use simulations of scalar mixing in non-reacting homogeneous isotropic decaying turbulence (HIT) such that (1) key modelling parameters can be taken from the direct numerical simulations without incurring additional modelling uncertainties and (2) direct validation is possible. Variants of Curl’s model are studied and direct comparison is sought with the variants’ performances in the context of standard (intensive) and sparse (such as MMC) particle approaches for the modelling of the probability density function (PDF). The second aim is to show the relative importance of micro-mixing and spatial diffusion in the presence of differential diffusion. The results demonstrate that MMC approximates the correct relaxation towards Gaussian independent of the mixing model’s variant. This is different from the standard PDF approach that requires a clear spatial localization given by the computational mesh to achieve a similar outcome. This spatial localization is not needed in MMC as the MMC mixing model already employs a localization in reference space. Differential diffusion effects can, however, only be accurately predicted if not only mixing but also spatial transport accounts for the differences in the molecular diffusion term. It is insufficient to adjust the mixing time scales only and future MMC models may require adjustments for accurate prediction capabilities.